[EPIC] native_datafusion scan improvements

[comphead]

What is the problem the feature request solves?

• native_datafusion doesn't report input metrics #3735
• [native_datafusion] Enable native_datafusion in auto scan mode — test failure tracking #3321
• Add native_datafusion V2 DataSource API reader #3481
• native_datafusion: no error thrown for schema mismatch when reading Parquet with incompatible types #3720
• [native_datafusion] [Spark SQL Tests] Schema incompatibility tests expect exceptions that native_datafusion handles gracefully #3311
• [native_datafusion] Subquery has not finished #2672
• [native_datafusion] Add support for reading Parquet field ids #3434
• [native_datafusion] Add support for reading row index metadata columns #3432
• Analysis of IgnoreCometNativeScan/IgnoreCometNativeDataFusion tests with native_datafusion (Spark 3.5.7) #3305